Parallel particle swarm optimization on chaotic solutions of dynamical systems

A parallel approach for finding chaotic solutions of general third-order autonomous dynamical systems with piecewise-linear vector field utilizing particle swarm optimization (PPSO) was proposed. PPSO is a metaheuristic method that can search very large spaces of candidate solutions using modern multi-core processors performance. To discover such sensitive systems in respect to chaotic behavior, special attention was devoted to the fitness function. Some of found systems are presented and analyzed in the article. A significant speedup using multi-core operations was proved.